Royal(ty) Pains: The Expansion Woes of Online Music Streaming Services

Online music streaming services have had a difficult year despite their increasing popularity. The problems services are experiencing south of the border may have an effect on how and when these services enter the Canadian market.

This year several high profile artists spoke out against, or withdrew from streaming services. In June, the members of Pink Floyd spoke out against Pandora’s lobbying attempts to decrease the royalties that streaming services pay to the artists in their libraries. In Early July, Thom York removed his music from Spotify. The Radiohead front man claims that the site’s business model does not properly compensate new artists. In late July, Aimee Mann filed a suit against MediaNet. The claims allege that MediaNet’s library includes a number of improperly licensed works and further claims that Mann has not received a royalty payment from the entity since 2006.


Before the industry can expand, streaming services must resolve their licensing problems. In 2011, ASCAP reached a U.S. licensing agreement with Spotify, granting the site access to over 8.5 million songs in the ASCAP catalogue. In that same year, however, over 200 indie labels withdrew their catalogs from several streaming services, including Spotify, Napster, Simfy and Rdio over concerns of under-compensation and for the fear that streaming would hurt sales for smaller artists.


Streaming services rely on artist participation and licensing agreements. These relationships must be stable in order for streaming to truly take off as an alternative to traditional radio, or services like iTunes. If artists do not have faith in the current business model, then it will be difficult for streaming services to expand. When popular artists publicly decry the system, it can sour potential relationships both with new users and artists who have yet to licence their music.


Some streaming services argue that the barrier to expansion is the cost of royalty payouts. Pandora scrapped a Canadian expansion in 2007 due to the cost of royalty payments in Canada. According to founder Tim Westergren, “the rates that have been proposed by the Canadian music rights societies are simply uneconomic.”  In addition, the current tariff rate may not be set until several years into the future. For example, one tariff on streaming services for the period from 2008-2010 did not receive certification until 2012. It may be difficult for a service to accurately gauge their viability in a new market if they may end up paying a substantially higher royalty rate than initially anticipated. Services may simply be unwilling to expand without certainty regarding these tariff rates.


While radio tariffs account for the fact that one play of a song can reach a large audience, on streaming services each play generally only reaches one user. Therefore the amount paid out for each play is much smaller. For example, one artist only received $1652.74 for over 1.5 million plays on Pandora. If these services increase in popularity, then perhaps the number of individual plays may eventually lead to payouts that resemble those for current radio rates. The argument from Pandora is that lower royalties now will mean long-term higher payouts. Lower royalties would allow services to pour more money into development and advertising. If services expand in size, then although the rate is lower the actual amount will increase due to the increased volume of traffic on the services.


Organizations that collect royalties for Canadian rights holders feel that their proposed rates represent fair market rates. In fact, in the U.S. over 125 artists signed an open letter opposing Pandora’s lobbying attempts to lower royalty rates. Moreover, despite Westergren’s assertions, Songza successfully negotiated a private agreement to licence music from Re: Sound and launched its “music concierge” playlist streaming service in Canada in 2012. Therefore there is potential for other services to come to an agreement in Canada, but the question remains whether royalties will present a problem for their bottom line.


Sadly, the string of artist complaints and withdrawals from several popular streaming services may temper online music streaming services expansion to Canadian markets, and the growth of the industry in general. Licensing agreements are a tremendous initial cost required to build a library expansive enough to attract users. If the relationship with artists remains unstable and services are unable to retain a diverse library, then attracting users to pay for these licences through subscriptions becomes a more difficult task. As a relatively small market, with the potential for retroactive tariffs, we may have to wait for the business model to further develop and grow before Canada enjoys the variety of services offered in other jurisdictions.

Allison McLean is an IPilogue Editor and a JD Candidate at Osgoode Hall Law School.

  1. I think it’s sad that it all comes down to money.. and that an industry that was so overpaid for so long is really going to complain about being paid appropriately now. I use torch music because it feeds off youtube, and I know the artists get their cut, but I’m really tired of the complaining. Things are changing. If they’re not up for it, they should get off the bus. There are plenty of new musicians that can accept how things are and make perfectly good music.

  2. Hey Allison,

    I think it is an interesting take on the situation that certain artists are withdrawing from semi-interactive webcasting services, like Pandora or Songza, due to perceived unfairness in compensation. I would suspect that is the exception rather than the rule (especially with respect to Mr. York’s activities concerning copyright), but certainly a trend worth monitoring.

    I’d like to add a couple points to this. Most of what I will say is summarized from a blog I wrote for IP Osgoode earlier the year when I attended the Copyright Board hearing concerning the tariff certification of Re:Sound for webcasting and semi-interactive webcasting.

    First, whether trends south of the border affect the Canadian Board’s tariff decision is quite literally the key issue. Re:Sound argued precisely that – the Board ought to take into account the US statutory hypothetical formula for royalty calculations of the “willing buyer/willing seller,” because modern technology has out-paced the current Canadian royalty structure with respect to services like Pandora et al. Not surprisingly, Pandora et al argued the opposite.

    However, while it is not my intention to suggest one formula is superior than the other, from a strictly legal perspective it is a very open question to what extent US royalty rates are probative in Canada. The significant constitutional and statutory differences which exist in that country (especially with respect to digital transmissions of music), have caused both the Canadian Board and the US Library of Congress to express a degree of scepticism of the relevance of their neighbour’s royalty structures.

    Second, the heart of this issue is whether semi-interactive webcasters ought to pay a royalty based on a per-play rate. At the Board, Re:Sound argued that the royalty structure ought to be a percentage of revenue + a per-play rate. Pandora et al argued for a royalty structured based primarily on some function of revenue.

    How do you structure a royalty for a service like Pandora – a “pull” technology that is free, ad-supported, with an option for paid subscription?

    You’re quite right to acknowledge the difficulty in comparing the royalty structures of semi-interactive webcasters with push technologies such as radio broadcasting, vis-a-vis one large audience at once v many individual listeners over time. However, radio broadcasters do not pay per-broadcast, as this blog suggests. It is easily calculable, but the royalty is not structured in this way. The royalty structure for commercial radio broadcasters is a function of: 1) amount of use of music, relative to air time, and 2) revenue, which is divided in tiers based on certain thresholds. The question then becomes – is receiving “only” x amount for 1.5 million streams on Pandora less than, equal to, or greater than, a radio broadcast that reaches 1.5 million listeners? Or, is that a relevant comparison at all?

    iTunes is mentioned, but iTunes is a pay-per-download service. It is much easier to structure a royalty based on a pay-per-download business model, because the costs are simpler to pass on. Youtube is a closer comparison in this respect.

    At the end of the day, in Canada a “balance” will have to be struck here. As you’ve clearly demonstrated, it is no secret that certain artists perceive that they are being undercompensated for their work. Certainly a lack of confidence in the system by creators is a bad thing. However, it is also no secret that for many users the feeling is quite mutual with respect to copyright tariffs. The real question is – what is the optimal royalty structure to appease both sides here? Per-play? Per-page impression? Per-unique visitor? Revenue? Or, maybe more creativity is needed.

Leave a reply

Your email address will not be published. Required fields are marked *

three × 3 =